🐕‍🦺 Leash Training: From Pulling to Perfect Walks

About this simulation

A dog on a walk is pulled toward interesting smells, squirrels and passers-by, while the leash sets a hard physical limit on how far it can wander from the handler. This scene models that tug-of-war from directly overhead: the handler follows a circular path, a wandering "distraction" pulls the dog outward, and the leash goes taut the instant the dog reaches the end of its slack.

How it works

• Leash tension — turns from green (loose) to red (taut) as the dog approaches leash length; a taut leash exerts a visible pull-back force.
• No correction — the handler keeps walking regardless of tension, so the dog learns pulling gets it closer to what it wants — pulling gets reinforced and worsens over time.
• Stop-and-Go — the handler freezes the instant the leash tightens and only resumes once the dog slackens the leash on its own, rewarding loose-leash position with forward motion.
• Loose-Leash + Treats — the handler occasionally drops a treat at heel height while the leash is loose, actively reinforcing the dog for staying close.
• Walk quality — a rolling score of how much time the leash spends loose versus taut under the selected method.

Controls

• Distraction drive — how strongly the dog wants to chase scents/squirrels (0–100).
• Leash length — the physical slack available, in metres (1–5m).
• Walking speed — the handler's base pace.
• Training method — switch between no correction, stop-and-go, and treat reinforcement to compare outcomes.
• ↺ Reset walk — restarts the walk and quality score. Drag to rotate, scroll to zoom.

Did you know?

Stop-and-go (also called "be a tree") works because it removes the dog's reward — forward movement toward the distraction — the instant it pulls, then restores that reward only for loose-leash position. Most dogs learn the pattern within a few dozen repetitions.